Fall 2019 Journal: Leveraging Low Carbon Fuel Standard Credits to Improve Electric Vehicle Rate Design

October 15, 2019 bppj Energy, Environment 0

This article is the third article featured in our Fall 2019 journal. For the complete journal, please see the “Journal Archive” tab above.

by Caroline Palmer

Edited by: Ben Menzies and Eddie Sun

The California Public Utilities Commission has an opportunity to reconsider how regulated electric utilities spend the proceeds from selling their Low Carbon Fuel Standard (LCFS) credits, given the program’s recent expansion. The Commission must also address the economically inefficient retail rates that deviate from the social marginal cost of producing electricity. This article proposes and analyzes two policy alternatives that would spend the expanded LCFS funds by advancing both electric vehicle (EV) adoption and economically efficient rates. I compare a fixed electric bill credit to a per-unit discount to existimg electric vehicle time-of-use rates. I find that the fixed credit is better than the per-unit discount across several criteria. I also recommend that California’s state agencies make a concerted effort to improve data collection and consumer engagement around reporting EV ownership and monitoring EV driving and charging. Effective policymaking is far more difficult without this information.

Introduction

As regulator of California’s investor-owned utilities, the California Public Utilities Commission (CPUC) ensures the safety and reliability of the state’s electric service. The commission also controls how the cost of that service is collected from customers as “just and reasonable” rates, aiming to protect ratepayers from monopolistic providers while also encouraging those providers to operate efficiently. This “rate-of-return” regulation involves setting the price that utility customers pay for electricity. The economies of scale in electricity service that maintain utilities’ marginal cost (MC) below their average cost have caused retail electricity rates in California (and many other states) to deviate substantially from the wholesale cost of electricity production, creating economic inefficiency.

Economically inefficient electricity rates happen for multiple reasons. In an efficient market, for every additional unit of electricity consumed, customers should be paying the marginal cost of producing that power (P = MC). However, they currently pay much more than that during most of the day to cover the fixed costs of the electricity grid, such as poles and wires. Those fixed costs must be paid, but rolling them into retail rates distorts the price signal that would otherwise inform customer decisions about how much power to consume at what time. Economists have long documented the inefficiency of setting retail prices that obscure marginal costs to society [1, 2, 3]. Most residential customers evidently cannot even perceive the marginal price of their own rates under the widely-used increasing block pricing scheme (in which customers are sorted into discrete tiers based on usage and higher-use tiers are charged a higher rate). Evidence suggests customers see the average of the price tiers rather than the price of each additional unit, complicating any potential customer response [4]. The CPUC must consider opportunities for making rates more transparent and much closer to the variable or marginal cost of electricity to consumers.

The CPUC has already taken a significant step in this direction with a 2015 decision. Residential customers will be switched by default onto time-of-use (TOU) pricing, meaning that they will face higher power prices at times of peak demand (for example between 4:00 – 9:00 p.m.). This can make rates more efficient because it addresses a separate inefficiency of the uniform, per-kilowatt-hour rate structure: the fact that marginal cost of generation and transmission varies dramatically by the time of day. An electricity supply curve reveals an increasing-cost industry in which high evening demand causes higher prices than does low nighttime demand. However, most residential customers never receive a price signal to change their behavior under an increasing block tier rate. With TOU rates, consumers will be able to perceive time differentiation and adjust their behavior accordingly. Although this change indicates the CPUC’s interest in bringing rates closer to MC, the TOU rates themselves will still remain well above marginal cost to collect the fixed costs of the grid, and more action is needed to bring rates closer to MC.

The CPUC must balance goals of reliable and efficient energy service with important statewide environmental priorities, such as increasing the use of electric vehicles (EVs). California’s policymakers have implemented measures that reduce fossil fuel combustion and carbon dioxide emissions to address climate change. This includes increasing the number of EVs owned and operated in the state. In 2018, Governor Jerry Brown signed an executive order setting a state goal of at least 5 million zero-emission vehicles on California roads by 2030, dramatically increasing the former 2025 target of 1.5 million vehicles [5]. The California Air Resources Board (CARB)’s recent Greenhouse Gas Inventory findings that California’s transportation sector is the largest source of emissions in the state underscores the urgency of such action. Emissions from transportation in fact increased 2 percent in 2016 [6].

The CPUC plays a key role in encouraging EV adoption by setting electricity rates (representing the cost of fuel for those vehicles) and regulating how investor-owned utilities spend the credit proceeds they earn through California’s Low Carbon Fuel Standard (LCFS). The LCFS is a state regulation requiring a 20 percent reduction in the carbon intensity of California’s transportation fuels by 2030. The mandate is accompanied by a market for tradable credits based on fuel source emissions reductions, in which producers of low-carbon fuels earn credits that producers of high-carbon fuels must buy to meet their emissions targets under the mandate.

Electric utilities may earn credits for electric vehicle charging within their territories as long as they “use all credit proceeds to benefit current or future EV customers” [7]. In 2014, the CPUC decided how electric utilities may spend those LCFS credit proceeds in order to boost EV utilization. The decision permits spending the money on reducing the upfront purchase cost of a vehicle or annually crediting EV customers’ utility bills [8]. Recently, the CARB expanded the incentives for using electricity as a transportation fuel, allowing utilities to also earn credits for building new charging infrastructure [9].

This paper frames the LCFS expansion and resulting IOU credits as an opportunity to of the economic problem of retail rate inefficiency. As electrification of transportation accelerates and the function of electric utilities evolves, the LCFS expansion creates a promising, politically-feasible occasion to address the inefficiencies in current rate design, by encouraging EV adoption through changes to the residential pricing structure for EV drivers. The lower the electricity bill for EV driving, the more incentive consumers have to switch to and to utilize EVs. If this can be achieved in a way that makes rates closer to marginal costs, it would be beneficial to both EV adoption and to efficiency in electricity consumption.

Alternatives

The policy alternatives considered in this analysis are intended to frame a conversation about the best possible rate scheme, given the limitations of rate-of-return regulation and the political and institutional inertia of the current rate structure.

While the most economically efficient outcome would be to charge all customers retail prices equal to the true social marginal cost (SMC) of electricity and assign a separate fee to ensure that the overall revenue requirement is met, such a radical change is not feasible in the short term. Instead, the LCFS expansion creates a window of opportunity for the CPUC to influence how utilities and ratepayers think about pricing. The CPUC can start with those ratepayers who drive EVs, since distribution utilities are mandated to focus LCFS funds on their benefit. Therefore, these alternatives offer options to utilize expanded LCFS funds in ways that specifically relate to EV retail rates:

  • Status Quo: Investor-owned utilities’ whole-house EV TOU rates
  • Alternative 1: Subsidize EV customer bills via per-kWh discount
  • Alternative 2: Subsidize EV customer bills via fixed monthly credits [10]

Status Quo: Investor-Owned Utilities’ Whole-House EV TOU Rates

Each of California’s three investor-owned utilities offers one or more special time-of-use rates for electric vehicle drivers. The LCFS policy directs the utilities to “provide rate options that encourage off-peak charging” [11] and these rates indeed set lower prices during periods of low demand. Even without targeted action by the CPUC to direct LCFS funds toward efficient retail rate remedies, it seems realistic to expect EV drivers to adopt special EV-TOU rates as TOU rates in general become mandatory across the state. Therefore, the alternatives presented in this analysis will be considered in comparison to EV-TOU rates rather than in comparison to California’s traditional increasing-block pricing methodology. As an example, Table 1 presents San Diego Gas & Electric’s (SDG&E) EV-TOU options.

Each of the IOUs offers the option to meter and bill EV electricity consumption separately; that is, to have different rate schedules for kilowatt-hours (kWh) consumed by the household and by the vehicle. This paper makes the practical assumption that most users will not have separate meters for their vehicle but will rather participate in a “whole-house” rate that treats all consumption according to the same rate schedule. From an economic standpoint, the whole-house rate makes sense because the cost of electricity at a given time is the same no matter how someone uses it. Therefore, power consumed by home appliances and power consumed by an EV battery ought to face the same price if consumed at the same time and on the same property.

EV drivers are unlikely to utilize a separate meter for financial reasons as well. Installing a new meter is costly and can require electrical updates. PG&E, for example, charges a $100 service fee for a second meter and estimates that a customer will have to spend an additional $2,000 to $8,000 to purchase the new technology [12].

Alternative 1: Subsidize EV Customer Bills Via per-kWh Discount

One way to incentivize EV adoption while addressing inefficiently high retail rates would be for utilities to use LCFS funds to offer customers a discount per kWh of consumption. To calculate the magnitude of the discount, utilities would divide their total LCFS credit proceeds by the total kWh consumed by EVs in their territory. Each kWh would be credited at that rate on the customer’s monthly bill if the customer is paying the EV rate. Whole-house EV-TOU rates, however, would obscure how much of the household load can truly be attributed to EVs – a problem, as subsidizing an entire household’s energy consumption does not clearly reward and incentivize EV ownership.

There could also be an egregious equity issue at hand: Applying this discount to an entire house would reward customers differently who consume different amounts of electricity that have nothing to do with their EV demand. Although consumption patterns can vary widely within income groups, lower-income EV users might generally have lower overall household electric consumption, meaning they would receive a smaller monthly discount than would larger consumers. Different allocations would end up being made between apartment-renters versus large homeowners or single versus multi-member households, even if their EV driving and charging behavior is similar.

Clearly, policymakers need to find a way to distinguish EV demand in order to avoid promoting a potentially regressive policy. CARB has issued guidance on how to estimate the consumption of non-metered residential vehicle charging. To do so, the agency recognizes that some EV users will inevitably use a separate meter for measuring EV charging. If this holds true, utilities are supposed to assume that the average daily consumption of non-metered EVs in their territories is equal to that of the metered EVs.13 CARB calculates the number of non-metered EVs in each service territory based on the California Vehicle Rebate Project database, and California Department of Motor Vehicles registration data. This information somewhat resolves the issue of how much whole-household consumption a utility can attribute specifically to the EV at that household. It would certainly also be helpful to utilities for ensuring that those EV owners register for one of the EV-specific rates as well.

Under time-varying rates, there is an opportunity to further calibrate the way a per-kWh discount is allocated, in order to better align with social marginal cost. The retail price in some EV-TOU periods may deviate more from the true average SMC in that period than in others. Off-peak retail rates might be further from off-peak SMC than on-peak rates are from the on-peak SMC. The CPUC could assign discounts proportionally to ensure that rates in each period be the same percentage higher than MC. Although this would require an extremely complicated methodology, it serves the purpose of more closely matching rates to marginal costs, and therefore it is the policy alternative considered here. By incorporating time-differentiated discounts, this proposal also notably evolves the CPUC’s 2014 rate reduction option and evaluation.

Alternative 2: Subsidize EV Customer Bills Via Fixed Monthly Credits

A different way to incentivize EV adoption and mitigate inefficiently high retail rates would be offering customers a fixed monthly bill credit using LCFS funds. Utilities would calculate the credit by dividing their total LCFS credits by the number of EV rate accounts. Unlike the per-kWh discount, this would not separate electricity utilized for EV charging, instead assigning a uniform credit across EV rate users. This methodology, therefore, would align with the assumption that most EV drivers do not separately meter residential EV consumption.
A bill credit could occur monthly or biannually, symmetric with the California Climate Credit. Utilities that already distribute climate dividends should not have difficulty allocating LCFS dividends to EV owners. A larger biannual credit may increase ratepayer awareness compared to a smaller monthly distribution, especially if the utility publicizes the larger, less frequent incentive. However, a case exists for selecting a monthly credit.

Like the per-kWh discount, the monthly bill credit could differ across customers. In the CPUC’s 2014 proceeding, PG&E suggested distributing an annual credit based on vehicle battery size, but General Motors challenged the idea that battery size accurately correlates with vehicle miles travelled and associated charging needs. A monthly credit could also vary across types of users, for example, based on consumption brackets of 0-1000kWh, 1001-2000kWh, and so on. However, I recommend rejecting this proposal since a core purpose of the incentive is rewarding each recipient for buying an EV, which may not relate to total household electric consumption.

Criteria

This paper considers three criteria when evaluating the stated policy alternatives:

  • Economic Efficiency
  • Impact on EV Adoption (Effectiveness)
  • Equity

Economic Efficiency

Interpreting the economic efficiency of each policy alternative requires determining its welfare cost as a function of how far rates under that policy deviate from the true social marginal cost (SMC). Therefore I first discuss why the existing EV-TOU rate structure is already quite far from the SMC.

It is difficult to precisely measure SMC for an entire state or even for a single utility service territory in California. This becomes particularly complicated when seeking values for the different periods under a time-varying rate. SMC changes frequently, by location, by time of day, and by season. The marginal cost of power can be approximated using the competitive wholesale price of electricity, which represents the variable fuel cost of generation. Friedman uses North American Electric Reliability Corporation data from 2009 to estimate the off-peak marginal cost of US electricity rates [13]. Friedman adds 13 percent% to the locational marginal price to approximate marginal ancillary services and distribution expenses, ultimately concluding that the April-June 2009 off-peak MC of power in California averaged 2.240 cents per kilowatt-hour [14].

More recently, Borenstein and Bushnell again estimate the social marginal cost of electricity across the country [15]. They find the average wholesale power price for California’s Independent System Operator control area to be 3.386 cents per kWh. After incorporating distribution line losses and omitting capacity costs and ancillary service costs, they calculate the average private marginal cost per kWh as falling between 3.2 and 4.6 cents. The authors also account for pollution externalities to determine the social marginal cost, calculating the hourly emissions damage of power generation regressed on load to identify the change in emissions in response to change in load. They conclude that the average social marginal cost per kWh in most parts of the state is under 6.5 cents.

Both of these marginal cost measurements indicate that California EV drivers on EV-TOU rates are paying well above the incremental cost of electricity. This is not surprising since the rates are explicitly designed to collect both fixed costs and marginal costs. Most of the SDG&E rates shown in Table 1 charge winter off-peak and on-peak prices over 15 cents higher than Borenstein and Bushnell’s average SMC, and SDG&E summer rates are even higher. Southern California Edison’s two-period EV-TOU rate, which charges 13 cents during off-peak and 37 cents during summer peak (24 cents during winter peak) is also clearly well above SMC.

One might expect the SDG&E rate, which offers customers a much lower super off-peak price in exchange for a $16 fixed cost (EV-TOU-5), to better approximate SMC. However, the super off-peak price is more than six cents higher than Friedman’s super off-peak estimate. From an efficiency standpoint, even the EV-TOU rate that comes closest to the theoretically optimal pricing structure is still deeply inefficient in terms of the price signal it sends to consumers. When consumers do not face accurate prices, they will not not allocate their consumption choices accordingly. Therefore, the status quo itself scores poorly on economic efficiency.

As for the policy alternatives, the per-kWh discount at first appears to be a great choice. Because the discount lowers the per-kWh rates, it brings them closer to marginal cost in the short run. Figure 1 demonstrates this efficiency gain. P0 represents the initial equilibrium price, or current electric retail rates. PLCFS represents the discounted per-kWh price, and PMC represents the SMC social marginal cost. Bringing rates down with a volumetric discount increases efficiency equivalent to the roughly triangular area below the demand curve between PLCFS and the vertical dotted line coming from P0. Bringing rates all the way down to marginal cost would also yield further economic efficiencies equivalent to the additional area under the demand curve.

While these short-run welfare gains are appealing, they only manifest for a certain portion of the ratepayers: those who own EVs and utilize EV rates. While lowering the rates for EVs serves the objective of bringing some rates closer to MC, the optimal efficiency gain needed in the electricity sector requires all rates to be closer to MC [16]. On principle, it is ill-advised to only achieve these gains for a single group – ill-advised enough to deter this analysis from recommending the per-kWh discount.

A fixed monthly credit, on the other hand, does not move rates away from the initial equilibrium price of P0. There are therefore none of the short-run efficiency gains from this alternative that there were from the per-kWh discount. In the long run, however, one would still expect efficiency gains from this policy alternative because it more closely approximates the idealized rate design that decomposes charges into fixed and variable elements. Receiving a monthly credit familiarizes customers and utilities with the idea that bills have a fixed component that could vary for different reasons. In fact, customers are already accepting the idea of a fixed bill charge by registering for SDG&E’s EV-TOU-5 option; 1,011 customers have already opted into this rate [17].

If the monthly credit policy alternative influences even more customers to start familiarizing themselves with fixed bill components, this could be an interim step to getting rates closer to MC, which will ultimately require a fixed charge. In this way, the monthly credit is a more economically efficient option than the per-kWh discount, and I accordingly rank the discount as a (-) and the bill credit as a (+).

Impact of EV Adoption

Judging the effectiveness of the policy alternatives requires determining whether more EVs will be bought and utilized based on whether kWh rates or overall monthly bills decline. The status quo EV-TOU rates are unlikely to cause EV adoption as they are so similar to normal default TOU rates. One might not expect much difference between a fixed credit and a per-kWh discount because they should both ultimately yield customers roughly the same savings. If EV owners all drove similar mileage and the per-kWh discount could accurately capture that utilization, then the total LCFS funds divided by kWh consumed as mileage would give the same savings when dispersed evenly across drivers as would a monthly bill credit equal to the LCFS funds divided by EV owner.

However, lower EV rates would theoretically encourage more EV charging than a fixed bill credit would, if consumers respond to marginal price signals. Increased EV charging achieves the stated LCFS purpose of reducing the carbon intensity of transportation fuels. The utilities have incentive to support this outcome, due to the increased electricity throughput. However, more electricity consumption has associated emissions; electricity may produce lower emissions than gasoline vehicle miles traveled, but policymakers may still be concerned about this outcome in a climate-conscious era.

From a consumer awareness perspective, neither a monthly credit nor a per-kWh discount is likely to stand out on an electric bill. Neither mechanism is very visible to consumers as just another line item on the utility bill; ratepayers will simply look at the total price and pay it. One benefit of the monthly bill credit option is that it is more tangible and easy to understand. A per-kWh discount is complicated for EV buyers to measure the magnitude of their credit. Understanding how much the credit will offset an EV purchase over the vehicle’s lifetime is important for the incentive to function properly. Even if the utility sums up the cumulative value of the discount every month to emphasize the value of the policy, it is certainly complicated for the consumer to understand.

Given the generally uncompelling effectiveness of either policy alternative, I rank the kWh discount as a (+), due to its potential to increase EV charging, and the fixed credit as neutral.

Equity

Incentivizing electric vehicle adoption and utilization generally raises equity concerns, as the costlier technology may be prohibitive to lower-income drivers. It would be reasonable to specifically target rebates toward residents in disadvantaged communities or toward those enrolled California’s discounted electric rates program. Since LCFS credits for EV charging are generated by existing EV drivers, the policy specifically directs the credits back toward those same drivers. Therefore, this section examines equity as it relates to the people who generated the LCFS credits.

Alternative 1, the kWh discount for whole-house metering, may not improve equity. As mentioned earlier, if the utility cannot reasonably isolate electricity consumed as vehicle fuel (which would be extremely difficult to achieve on the whole-house meter), it would subsidize other household consumption in addition to EV charging, thereby allocating LCFS credit proceeds away from their mandated purpose. Referring again to the EV uptake rates in Figure 2, the 12,160 SDG&E customers who have a single meter clearly outnumber the 364 customers who have a separate EV meter. This discrepancy means that the per-kWh discount, even if it utilized the CARB’s recommended methodology for estimating non-metered EV usage, will be based on a very small sample size and will likely be distorted. This alternative suddenly appears very similar to a fixed credit due to the difficulty of precisely measuring EV consumption as it is simply a value averaged across most EV users.

A fixed credit may have the opposite problem, in that all customers will receive the same credit regardless of how much they utilize an EV. This may fail to reward increases in consumption of electricity as transportation fuel, a goal of the LCFS program. Given that the per-kWh discount effectively becomes a fixed credit after averaging across the non-metered EV users, however, I choose to rank both policy alternatives as a (-).

Conclusion

The CPUC currently has the opportunity to rethink the way utilities spend the proceeds from their LCFS credits. Given the economic inefficiency of retail rates set far above the SMC social marginal cost, the CPUC must also focus on the long-term goal of bringing all rates closer to SMC. I propose a set of policy alternatives in which the agency can address both objectives at once.

Below, I summarize the qualitative rankings of the two policy alternatives according to the three criteria analyzed in this paper. All criteria were weighed equally; efficiency was not given special treatment given that it is not as prominent a goal of the LCFS as is EV adoption.

Ultimately, I recommend the fixed credit over the per-kWh discount. While the volumetric discount is an expedient way of getting EV rates closer to MC, it is not the efficient principled to improve just one group of electric rates when the overarching goal must be to get all rates closer to SMC. The fixed credit, on the other hand, could acclimate customers to bills with fixed components, which could pave the way for the future fixed costs that would necessarily accompany marginal-cost prices. None of the other criteria strongly influences the decision for one policy alternative over the other except for the potential effectiveness of the per-kWh credit to increase EV charging and, therefore, EV utilization.

In addition to this recommended policy, some additional improvements will strengthen the nexus between electric vehicle adoption and efficient electricity rates are necessary for successful policymaking. Fundamental assumptions of this paper include that EV drivers are using EV rates effectively; that they know about those rates, that they have chosen to enroll in them, and that the utilities accurately track EVs in their service territories due to the rate utilization. The 2014 CPUC proceeding identified that EV drivers are often not on these rates and that utilities may not know that non-metered EVs even exist unless a customer notifies the utility when requesting a rebate. The utilities and agencies involved in LCFS must work with the California Department of Motor Vehicles DMV to identify non-metered vehicles. Even if the utilities better track the number of EVs in their territory, determining charging behavior of individual EV drivers is still a challenge. This limits the per-kWh discount policy alternative, and it is critical for measuring LCFS credits accurately in general. In addition to recommending that the Commission adopt the monthly bill credit policy alternative, I also strongly recommend that the state dedicate resources to tracking EV ownership and usage in California. Improved data will allow agencies and utilities to incentivize EV uptake across the state more effectively.

___________________________________________________________________________

Caroline Palmer (MPP ’19) is an alumna of the Goldman School of Public Policy.

Endnotes

  1. Borenstein, Severin (2005). “The Long-Run Efficiency of Real-Time Electricity Pricing”. The Energy Journal, V26:N3, pp. 93-116.
  2. Friedman, Lee S. (2012). “Consumer-Friendly and Environmentally-Sound Electricity Rates for the Twenty-First Century,” Goldman School of Public Policy.
  3. Allcott, Hunt (2009). “Real time pricing and electricity markets”. Harvard University.
  4. Borenstein, Severin and Bushnell, James (2018). “Do Two Electricity Pricing Wrongs Make a Right? Cost Recovery, Externalities, and Efficiency”. National Bureau of Economic Research, Working Paper No. 24756.
  5. Office of Governor Brown (2018) “Governor Brown Takes Action to Increase Zero-Emission Vehicles, Fund New Climate Investments”. https://www.gov.ca.gov/2018/01/26/governor-brown-takes-action-to-increase-zero-emission-vehicles-fund-new-climate-investments/. Accessed 8 Dec. 2018.
  6. California Air Resources Board (2018a). “California Greenhouse Gas Emission Inventory – 2018 Edition”.
  7. California Air Resources Board (2015). “Low Carbon Fuel Standard: Final Regulation Order”. Available at: https://www.arb.ca.gov/fuels/lcfs/fro_oal_approved_clean_unofficial_010919.pdf.
  8. CPUC 2014
  9. California Air Resources Board (2018b). “Press Release: CARB amends Low Carbon Fuel Standard for wider impact”. https://ww2.arb.ca.gov/news/carb-amends-low-carbon-fuel-standard-wider-impact. Accessed 8 Dec. 2018.
  10. Notably absent from this list of methods to incentivize EV adoption is the idea of upfront purchase rebates. Providing rebates would be a very reasonable alternative for spending the LCFS proceeds, as the CPUC decided in 2014. An upfront rebate is especially appealing with regard to the goal of EV uptake, as it addresses the bounded rationality of people who would benefit from EV ownership in the long run but are stopped by the “sticker shock” of the capital cost. However, as this paper focuses on using LCFS funds to increase efficiency in rate design as a primary goal, rebates have not been included as an alternative here.
  11. California Air Resources Board (2015). “Low Carbon Fuel Standard: Final Regulation Order”. Available at: https://www.arb.ca.gov/fuels/lcfs/fro_oal_approved_clean_unofficial_010919.pdf.
  12. Pacific Gas & Electric (2018). “Install a Level 2 charging station for your electric vehicle”. https://www.pge.com/en_US/residential/solar-and-vehicles/options/clean-vehicles/electric/charger-installation.page.
  13. California Air Resources Board (2018c). “Notice of Upcoming Low Carbon Fuel Standard Credits Release for Non-Metered Residential Electric Vehicle Charging”.
  14. Friedman, Lee S. (2011). “The importance of marginal cost electricity pricing to the success of greenhouse gas reduction programs,” Energy Policy, V39, pp. 7347-7360.
  15. Borenstein, Severin and Bushnell, James (2018). “Do Two Electricity Pricing Wrongs Make a Right? Cost Recovery, Externalities, and Efficiency”. National Bureau of Economic Research, Working Paper No. 24756.
  16. In general, it is also not prudent to distort groups of rates because if they were at MC, they should stay there as representative of true incremental costs. One must assume that the CPUC has already done the best it can within the institutional constraints of cost-of-service regulation to ensure that rates are as close as possible to MC, even if they clearly are not actually very close at all.
  17. San Diego Gas & Electric (2018). “San Diego Gas & Electric Company (U 902-E) Quarterly Report on the Progress of Residential Rate Reform (PRRR)”